AI Agent

What your AI knows, remembers, and how it behaves.

Memory Crystals

Persistent context across every session

Save architectural decisions, code patterns, project conventions, and domain knowledge as named memory. Your AI loads them at the start of every session — it knows your codebase before you type a word.

Named memories — create, edit, and organize by project
Automatically loaded into agent context at session start
Cloud-synced across devices when signed in
Works fully offline with local storage fallback

Skills

Extend what your AI agent can do. Skills are modular capability packages — each one teaches the agent a new workflow, tool integration, or domain-specific behavior. The agent itself can write and register new skills on the fly.

Build your own — or ask the agent to create one for you
Skills compose — combine multiple for complex workflows
Stored locally in your project alongside your code

Rules

Define how your AI behaves at the project, folder, or file level. Rules are persistent instructions that shape every response — coding standards, naming conventions, architectural constraints.

Scoped rules — global, workspace, or file-pattern level
Version-controlled alongside your code
Automatically injected into agent context

Project Blueprint

Auto-generated architecture documentation for your codebase. The blueprint gives your AI deep structural awareness — file relationships, dependency graphs, and entry points.

Auto-maps your project structure on open
Feeds into agent context for smarter suggestions
Exportable as documentation

Regen

Record any agent task as a replayable workflow. Hit Start Recording, ask the agent to do anything — refactor code, scaffold a feature, generate a component — then save it with a name. The exact steps are captured and can be replayed on any project, with variables you define.

Smart mode — adapts the workflow to the new project context
Instant mode — exact replay, same steps every time
Works across projects — save once, reuse anywhere
Workspace

Visual, navigable, designed around how you actually work.

Constellation View

Visual map of your entire project

See your project as a living graph — files as nodes, dependencies as edges, and the AI's attention as glow. Navigate complex codebases by structure instead of scrolling through file trees.

Interactive node graph of files and dependencies
Click any node to open the file instantly
See which files the agent is reading in real time
Highlights clusters, orphans, and hot paths

Built-in Browser

Preview and test web applications directly inside the IDE. No alt-tabbing, no context switching. The agent can see what you see — and debug it alongside you.

Embedded Chromium browser panel
Hot-reloads alongside your dev server
Agent-aware — reference what's on screen in chat

Custom Project Colors

Assign a unique color to each project. The title bar, sidebar, and status bar shift to that color — so you always know which project you're in at a glance.

Per-project color themes — set once, always visible
Instant visual identification across windows
Prevents committing to the wrong repo

Priority Machine

Intelligent task prioritization

persīv surfaces what matters most. The Priority Machine ranks your open tasks, files, and agent suggestions by relevance, recency, and impact — so you always know what to work on next.

Ranks tasks by project context and recent activity
Surfaces related files when you start working
Reduces decision fatigue — fewer choices, better flow
Team & Business

Visibility, control, and governance across your organization.

Dashboard

A web portal for managing your org, members, API keys, budgets, and billing. Accessible from any browser — everything you need to administer your persīv deployment in one place.

Org management — members, roles, invitations
Subscription and billing portal
Audit logs and usage analytics

Cost Tracking

Token-level spend attribution by model, user, and project. Updated live as requests flow. No waiting for a monthly invoice to find out what you spent.

Per-request cost breakdown in the IDE
Org-wide analytics in the dashboard
Budget estimates based on public model pricing

Key Sync

Set your API keys once in the dashboard. They're securely distributed to every member of your org — encrypted at rest, never visible in plaintext, and instantly revocable.

Org-wide key management from the dashboard
Keys encrypted at rest — never exposed to users
Individual keys also supported for personal use
$

Simple, Honest Plans

Two tiers. No usage-based billing. No metered API calls. Free gets you 20 cloud messages a day plus unlimited local model. Pro at $9.99/seat/mo unlocks everything.

Free
20 cloud messages/day
Unlimited Looking Glass local model
BYOK support, memory crystals, cost tracking
Pro — $9.99/seat/mo
Unlimited cloud messages
Team org, shared keys, budget controls
Full audit trail, analytics, priority support

Budget Controls

Set monthly spend ceilings at the org, team, or individual level. When the threshold is hit, requests stop. Estimates based on token counts and public model pricing.

Hard caps enforced in the IDE
Granular — org, team, or individual

Ready to start?

All of this. Free to download.

Every feature on this page ships in the free tier — except team features, which require Pro. Download persīv and start building.

Download Free View Pricing →